ranking stuff

ranks-and-stuff.R

rank_limited <- function(x, method = c("auto", "shell", "quick", "radix")) {
  u <- sort(unique(x), method = method)
  match(x, u)
}

rank_limited2 <- function(x, method = c("auto", "shell", "quick", "radix")) {
  u <- unique(sort(x, method = method))
  match(x, u)
}

x <- round(runif(1e4), 4)
res <- bench::mark(
  rank_average   = rank(x, ties.method = "average"),
  rank_first     = rank(x, ties.method = "first"),
  rank_last      = rank(x, ties.method = "last"),
  rank_min       = rank(x, ties.method = "min"),
  rank_max       = rank(x, ties.method = "max"),
  frank_average  = data.table::frank(x, ties.method = "average"),
  frank_first    = data.table::frank(x, ties.method = "first"),
  frank_last     = data.table::frank(x, ties.method = "last"),
  frank_random   = data.table::frank(x, ties.method = "random"),
  frank_max      = data.table::frank(x, ties.method = "max"),
  frank_min      = data.table::frank(x, ties.method = "min"),
  frank_dense    = data.table::frank(x, ties.method = "dense"),
  order_auto     = order(x, method = "auto"),
  order_shell    = order(x, method = "shell"),
  order_radix    = order(x, method = "radix"),
  limited_auto   = rank_limited(x, method = "auto"),
  limited_shell  = rank_limited(x, method = "shell"),
  limited_quick  = rank_limited(x, method = "quick"),
  limited_radix  = rank_limited(x, method = "radix"),
  limited2_auto  = rank_limited2(x, method = "auto"),
  limited2_shell = rank_limited2(x, method = "shell"),
  limited2_quick = rank_limited2(x, method = "quick"),
  limited2_radix = rank_limited2(x, method = "radix"),
  check = FALSE
)

print(res, n = Inf)
#> # A tibble: 23 x 13
#>    expression          min   median `itr/sec` mem_alloc `gc/sec` n_itr  n_gc
#>    <bch:expr>     <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl> <int> <dbl>
#>  1 rank_average     1.04ms   1.24ms      767.  421.96KB     4.06   378     2
#>  2 rank_first      428.3us  508.7us     1843.  381.75KB    12.6    879     6
#>  3 rank_last       443.4us  510.2us     1840.  430.16KB    15.1    852     7
#>  4 rank_min         1.01ms   1.15ms      817.  351.95KB     6.19   396     3
#>  5 rank_max         1.01ms   1.16ms      811.  351.95KB     4.06   399     2
#>  6 frank_average    1.25ms   1.35ms      705.    2.06MB     4.05   348     2
#>  7 frank_first      1.54ms   1.64ms      587.  119.06KB     0      294     0
#>  8 frank_last       1.26ms   1.38ms      689.  143.78KB     4.07   339     2
#>  9 frank_random     2.27ms   2.49ms      388.  219.57KB     0      194     0
#> 10 frank_max        1.26ms   1.42ms      645.  143.78KB     2.02   320     1
#> 11 frank_min        1.26ms   1.41ms      670.  143.78KB     2.01   333     1
#> 12 frank_dense      1.25ms   1.44ms      638.  143.78KB     2.02   316     1
#> 13 order_auto      316.3us  363.6us     2506.   39.11KB     2.02  1243     1
#> 14 order_shell     908.7us   1.04ms      917.   39.11KB     2.02   455     1
#> 15 order_radix     308.8us  357.7us     2559.   39.11KB     2.02  1269     1
#> 16 limited_auto    708.7us  845.6us     1108.   524.2KB    10.4    531     5
#> 17 limited_shell   824.5us 956.55us      977.  546.61KB    10.4    468     5
#> 18 limited_quick   754.9us  904.1us     1041.  521.43KB    10.4    499     5
#> 19 limited_radix   712.4us    835us     1140.  521.43KB    10.4    550     5
#> 20 limited2_auto   648.8us  786.9us     1186.  436.53KB    10.4    568     5
#> 21 limited2_shell    908us   1.05ms      891.  447.94KB     6.15   435     3
#> 22 limited2_quick  799.2us 914.85us     1043.  436.53KB    10.4    500     5
#> 23 limited2_radix  646.4us  789.3us     1202.  436.53KB     8.26   582     4
#> # ... with 5 more variables: total_time <bch:tm>, result <list>, memory <list>,
#> #   time <list>, gc <list>
res <- res[order(res$`itr/sec`), ]
res$expression <- mark::fact(as.character(res$expression))
mark::flip(tail(res))
#> # A tibble: 6 x 6
#>   expression          min   median `itr/sec` mem_alloc `gc/sec`
#>   <fct>          <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 order_radix       309us    358us     2559.    39.1KB     2.02
#> 2 order_auto        316us    364us     2506.    39.1KB     2.02
#> 3 rank_first        428us    509us     1843.   381.8KB    12.6 
#> 4 rank_last         443us    510us     1840.   430.2KB    15.1 
#> 5 limited2_radix    646us    789us     1202.   436.5KB     8.26
#> 6 limited2_auto     649us    787us     1186.   436.5KB    10.4
ggplot2::autoplot(res)